Accuracy Analysis of Robot Binocular Vision Calibration

Abstract

The accuracy of calibration is a critical link which affects the accuracy of the machine vision system. Several factors affecting the accuracy of the binocular vision calibration were analyzed. To construct a reprojection error was as a cost function. Deference controllable factors were set to getting intrinsic parameters. And then reprojection error was depicted utilizing a variety of graphical form, such as 2D coordinate drawings, line charts and cross-hatched map. The correlation between parameters which were paid attention in calibration process was analyzed clearly and directly. The selection of optional parameters is described based visualization. The selection basis of optional parameters was offered by experiments.

Drilling errors compensation of industrial robot based on 3D passive scale tracker

“…Nubiola et al [14][15][16] compensates the pose error of the robot by means of kinematic calibration and neural network using the vision measurement system. Additionally, Shu et al [17][18][19] improve the accuracy of the vision measurement system by reducing the measurement noise and optimizing the selection of feature points of the measured object, and measure and compensate the robot pose errors in real-time. Drawstring displacement sensors and ballbars are capable of measuring motion errors in a single direction.…”